What are the general rules / guidelines when doing a machine capability study?

M

maxschnau

WHat are the general rules / guidelines when doing say a 30 pc measurement capability study on a dimension on a part. Say one part is out of spec, can you get another part?, can you refixture the part?, etc, and if you have the anser to these rules, what source do they come from?

THANKS!!!
 

Jim Wynne

Leader
Admin
maxschnau said:
WHat are the general rules / guidelines when doing say a 30 pc measurement capability study on a dimension on a part. Say one part is out of spec, can you get another part?, can you refixture the part?, etc, and if you have the anser to these rules, what source do they come from?

THANKS!!!
Please explain what you mean by "capability study." What information are you hoping to get from the study?
 
M

maxschnau

I mean capability study - Initial Capability Studay as is required by Automotive. I would be looking to achieve a 1.33 or above CPK.
 

Jim Wynne

Leader
Admin
maxschnau said:
I mean capability study - Initial Capability Studay as is required by Automotive. I would be looking to achieve a 1.33 or above CPK.
You can't do that (by default) with just 30 pieces. Here are the default requirements as stated in the AIAG PPAP manual (3rd Edition):
For those characteristics that can be studied using X-bar and R charts, a short term study should be based on a minimum of 25 subgroups containing at least 100 readings from consecutive parts of the significant production run...

Furthermore, you can't calculate Cpk without first verifying that the process is in a state of statistical control, and I don't know how you can do that with a 30-piece sample.
 
M

maxschnau

OK, Let's say it's a 300 Piece Study. What are the rules of remeasuring or refixturing or replacing a part that is considered a flyer by the data.........
 
R

ralphsulser

Plus, when you find something in your sample size out of spec. you can't throw it out and ignore it, or change the gauge either :rolleyes:
 
M

maxschnau

OK, Let's say it's a 300 Piece Study. What are the rules of remeasuring or refixturing or replacing a part that is considered a flyer by the data.........
 

Miner

Forum Moderator
Leader
Admin
maxschnau said:
OK, Let's say it's a 300 Piece Study. What are the rules of remeasuring or refixturing or replacing a part that is considered a flyer by the data.........

You cannot ignore a part that is flyer unless there is a legitimate reason. By your questions about refixturing, it appears that you suspect that the part was incorrectly loaded into a measurement fixture, or there was debris between the part and fixture that threw the reading out. In a case such as these, you could legitimately clean the fixture, remeasure the same part and substitute the results.

However, in another example, this would not be true. Let's say that during the manufacture of product, a change occurred (material lot, etc.), the process shifted, then was adjusted back. In this example, the "flyer" is a legitimate part of the data that is telling you that the process is not always in control. You cannot simply throw this data out. You have to bring the process into control and re-study.
 
M

maxschnau

MIner, thanks for the explanation - it was exactly what I was looking for.
 
Top Bottom